摘要
脱线签名的验证较难,他仅依靠签名图像的静态信息,而书写过程中的动态信息几乎完全消失。针对脱线手写签名识别的特点,提出基于提升小波变换的特征选取方法,将传统的结构特征与统计特征有机结合起来。运用K-L变换对已提取的特征向量进行降维。最后通过支持向量机进行真伪识别。实验结果表明该算法对测试样本具有高识别率。
Identification of off- line handwritten signature is hard, because it only depends on static information,dynamic information is disappears. Aimed at the recognition characteristic of off -line handwritten signature, the feature selection method based on lifting wavelet transform is presented,and structure feature is combined with statistical feature both are conventional method. Extracted eigenvector is compressed by K - L transform. At last, true signature and forge signature are distinguished through support vector machines. The experiment results confirm given algorithm can reach satisfied identical rate for testing stylebook.
出处
《现代电子技术》
2007年第10期97-99,共3页
Modern Electronics Technique
基金
教育部回国人员科研启动资金([2004]527)
西南交通大学校基金(2005A02)
